Detecting-Alzheimer-s-Disease-in-MRI-using-Deep-Learning
Alzheimer's disease (AD) is a progressive disorder that causes difficulty in language and decision-making, memory loss and a decline in other cognitive abilities. Since there is no definite cure for this, diagnosing the disease in an early stage is the only prevention. However, detecting this disease in its early stages is one of the most difficult and challenging aspects as the symptoms of the early stages may start very early in life and are very subtle. Hence, there are instances where the disease isn’t caught until it is in its final stages. There have been factors found like genetics, age and lifestyle choice that can contribute to and even increase the chance of developing the disease. It's also critical to understand that detecting and treating AD can ease symptoms and enhance the patient's overall quality of life. In this study, our main objective is to detect if a person has AD using Magnetic Resonance Images (MRIs) and at which stage it is by exploring deep learning applications. One more objective of our study is to find out if changing the color scheme of MRIs has any effect on the performance of the initial goal. The present study successfully accomplished its objectives, as evidenced by the high accuracy of the proposed approaches in addressing the problem statement.
The images in the dataset are already preprocessed which makes it easier and faster to work with. The frequency of images of each category are: Class-1: 896, Class-2: 64, Class-3: 3200 and Class-4: 2240.
Distribution of images in the dataset
Sample Dataset-
*This project was published as a paper at the International Conference of Internet of Things 2023